← AI Hashrate 中文

Ryzen AI Max+ 395 128GB vs RTX 3080 10GB for Local LLMs

The RTX 3080 10GB has about 253% more memory bandwidth (760 vs 215 GB/s). Batch-1 decode tracks bandwidth closely, so a model that fits both cards should decode roughly 253% faster on the RTX 3080 10GB. The Ryzen AI Max+ 395 128GB fits more of the catalog at Q4/8K — 51 vs 17 of 69 models (96 vs 10 GB VRAM). Largest model that fits the Ryzen AI Max+ 395 128GB but not the RTX 3080 10GB: DeepSeek-V4-Flash (158.1B). MSRP is $2,000 for the Ryzen AI Max+ 395 128GB vs $699 for the RTX 3080 10GB. MSRP is a launch list price, not live retail — Amazon prices move; click through for the current price. In the Q4/8K head-to-head below (39 shared models), the RTX 3080 10GB posts the higher tok/s in 39 rows vs 0. Methodology.

Spec comparison

Ryzen AI Max+ 395 128GBRTX 3080 10GB
VendorAMDNVIDIA
VRAM96 GB10 GB
Memory typeLPDDR5XGDDR6X
Memory bandwidth215 GB/s760 GB/s
TDP120 W320 W
MSRP (list)$2,000$699
BuySearch Amazon (affiliate)Search Amazon (affiliate)

MSRP is the launch list price, not live retail. Buy links are Amazon affiliate links — the price after you click is the live price.

Q4 / 8K head-to-head — top 15 of 39 shared models

ModelParams (B)Ryzen AI Max+ 395 128GB tok/sRyzen AI Max+ 395 128GB fitsRTX 3080 10GB tok/sRTX 3080 10GB fitsWinner
MiniCPM5-1B1.08126.7 est.Yes447.8 est.YesRTX 3080 10GB
Qwen3.5-2B2.068.4 est.Yes241.8 est.YesRTX 3080 10GB
DeepSeek-R1-0528-Qwen3-8B-layer-mix-bpw-3.8-mlx2.359.5 est.Yes210.3 est.YesRTX 3080 10GB
gpt-oss-20b21.538.0 est.Yes208.0 est.NoRTX 3080 10GB
Llama-3.2-3B-Instruct3.242.8 est.Yes151.1 est.YesRTX 3080 10GB
Nemotron-3-Nano-Omni-30B-A3B-Reasoning-BF1630.045.6 est.Yes149.1 est.NoRTX 3080 10GB
gemma-4-26B-A4B-it25.236.0 est.Yes148.3 est.NoRTX 3080 10GB
Qwen3-30B-A3B-Instruct-250730.545.6 est.Yes135.7 est.NoRTX 3080 10GB
GLM-4.7-Flash31.245.6 est.Yes130.2 est.NoRTX 3080 10GB
Qwen3-4B-Instruct-25074.034.2 est.Yes120.9 est.YesRTX 3080 10GB
NVIDIA-Nemotron-3-Nano-4B-BF164.034.2 est.Yes120.9 est.YesRTX 3080 10GB
Qwen3-4B-Thinking-25074.034.2 est.Yes120.9 est.YesRTX 3080 10GB
Qwen3.5-35B-A3B34.745.6 est.Yes112.8 est.NoRTX 3080 10GB
qwen35b-a3b-fable-sft-abliterated34.745.6 est.Yes107.1 est.NoRTX 3080 10GB
Ornith-1.0-35B35.045.6 est.Yes105.4 est.NoRTX 3080 10GB

est. = bandwidth-based estimate; measured = curated real benchmark that overrides the estimate. Fits = weights + KV(8K) + 1 GB overhead ≤ 95% of VRAM.

Fit coverage at Q4 / 8K

Ryzen AI Max+ 395 128GBRTX 3080 10GB
Catalog models that fit51 of 6917 of 69
Largest exclusive fitDeepSeek-V4-Flash (158.1B)—

Ryzen AI Max+ 395 128GB details → · RTX 3080 10GB details →